Estimating three-dimensional nonlinear panel data models with interactive effects

This article considers two three-step estimation approaches to three-dimensional nonlinear panel data models with interactive effects correlated with the independent variables. We propose a Nonlinear Least Squares-Principal Component Analysis-Grid Search iterated approach to estimate the static model, and a GMM-PCA-GS iterated method to the dynamic model. Monte Carlo experiments examine the finite sample performance of the methods.     

A panel data model for subjective information on household income growth

Subjective expectations about future income changes are analyzed, using household panel data. The models used are extensions of existing binary choice panel data models to the case of ordered response. We consider static models with random and fixed individual effects. We also look at a dynamic random effects model which includes a measure for permanent and transitory income. We find that income change expectations strongly depend on realized income changes in the past: those whose income fell, are more pessimistic than others, while those whose income rose are more optimistic. Expected income changes are also significantly affected by employment status, family composition, permanent income, and past expectations. Expectations are then compared to the head of household’s ex post perception of the realized income change for the same period. The main finding is that rational expectations are rejected, and that in particular, households whose income has decreased in the past underestimate their future income growth.     

Nonlinear models with panel data

Panel data play an important role in empirical economics. With panel data one can answer questions about microeconomic dynamic behavior that could not be answered with cross sectional data. Panel data techniques are also useful for analyzing cross sectional data with grouping. This paper discusses some issues related to specification and estimation of nonlinear models using panel data.JEL Classification: C230The research behind this paper was supported by the National Science Foundation, the Gregory C. Chow Econometric Research Program at Princeton University, and Danish National Research Foundation (through CAM at the University of Copenhagen). The author thanks Ekaterini Kyriazidou, Hong Li, Marina Sallustro, and the editors for helpful suggestions.     

Trade effects of regional integration: A panel cointegration analysis

The paper quantifies the most likely trade effects of the exceptional cases of the GATT/WTO system, namely, Regional Integration Agreements, on the selected member as well as non-member countries of the EU, NAFTA, MERCOSUR and AFTA. To this end, the gravity model was estimated through fixed effects model and panel cointegration analysis. It was found that the explanatory power of the latter has superseded the former one. For the case of EU, it was found that the intra-union trade-creation effect is approximately six times larger than extra-union effects. In NAFTA, exports to outside countries are significantly diverted. For MERCOSUR, on the other hand, results indicate that the integration has not contributed to intra-union trade. The members are still significantly dependent on extra-union imports, just like the members of AFTA.     

A note on S in a spatially correlated error components regression model for panel data

The ordinary least squares based estimator of the disturbance variance in a panel regression model with spatially correlated error component is shown to be asymptotically unbiased and weakly consistent without any restrictions on the regressor matrix.     

Detecting multiple factors in panel data: an application on the growth of local regions in China

Due to unbalanced growth in China’s local regions, we construct a panel data model with multiple common factors to examine the differences among the growth factors in these areas. This article shows the various impacts from the supply and demand sides on economic growth. Different from the demand side, the supply-side impacts have permanent influences. This article focuses on these deep and profound impacts to explain the reasons behind China’s fast economic growing. By using data on 27 regions from 1958 to 2013, we summarize the main permanent influences along three lines. The first comes from the coastal regions, which have learned modern technology and systems from foreign companies, such as in Guangdong, Zhejiang, Fujian and Liaoning. The second comes from big cities, such as Beijing and Shanghai, in which a huge migration has given the companies opportunities to recruit excellent workers, making the resource allocation specialized and more efficient. The third is from the government’s major public works, which have improved areas’ infrastructure and assisted long-run economic growth, such as for Sichuan, Guangxi and Yunnan.     

The proper panel econometric specification of the gravity equation: A three-way model with bilateral interaction effects

We argue that the proper specification of a panel gravity model should include main (exporter, importer, and time) as well as time invariant exporter-by-importer (bilateral) interaction effects. In a panel of 11 APEC countries, the latter are highly significant and account for the largest part of variation. First version received: February 2001/Final version received: June 2002 RID="*" ID="*"  We are grateful to two anonymous referees and Robert Kunst for their helpful comments.     

Inventory investment and financial constraints in the Italian manufacturing industry: A panel data GMM approach

Three large unbalanced panels of Italian manufacturing firms observed over the period 1991–2009 are employed to assess, by means of a dynamic GMM approach, whether the existence of financial frictions is suitable to explain deviations of inventories from their long-run path. A negative response of inventory investment to the presence of financial burdens might provide evidence of a significant role played by the financial framework in conditioning the real side of the economy, especially during recession years, when liquidity problems arise. The negative effect is found over the entire analyzed period, with firms' dimensional aspects accounting more than risk characteristics to explain the phenomenon, but the inclusion of recessionary dummies into the model leads to controversial and puzzling results. A significant recessionary effect is found during the Nineties, accounting for inventories being more sensitive to financial frictions during the main recessionary peaks, 1993 and 1996. The result is not confirmed by the most recent estimates, especially the ones referring to the 2008–2009 recessionary shock, whose effects are investigated for the first time by a paper addressing the inventory investment–financial constraints subject. Alternative hypothesis for the proposed results have been tested on data. Firms were found to rely on inventory decumulation to a lesser extent compared to the past, to generate internal financing. More specifically, disinvestments in financial assets were found to represent, as a matter of fact, one of the main drivers adopted to ease liquidity tensions: a negative and strongly significant relationship with inventory investment was detected, after controlling for short-run liquidity constraints at firm level. By contrast, only a weak negative relationship was established with fixed capital during the same recessionary biennium.     

Three-dimensional panel data models with interactive effects: Estimation and simulation

This paper considers three-dimensional panel data models with unobservable multiple interactive effects, which are potentially correlated with the covariates. We propose an iterated OLS (IOLS) approach for the estimation of the static model and an iterated GMM (IGMM) method for the dynamic model. Monte Carlo simulations illustrate that our methods perform well in finite samples.     

Estimation of technical inefficiency effects using panel data and doubly heteroscedastic stochastic production frontiers

In previous studies, measures of technical inefficiency effects derived from stochastic production frontiers have been estimated from residuals which are sensitive to specification errors. This study corrects for this inaccuracy by extending the doubly heteroscedastic stochastic cost frontier suggested by Hadri (1999) to the model for technical inefficiency effects. This model is a stochastic frontier production function for panel data as proposed by Battese and Coelli (1995). The study uses, for illustration of the techniques, data on 101 mainly cereal farms in England. We find that the correction for heteroscedasticity is supported by the data. Both point estimates and confidence intervals for technical efficiencies are provided. The confidence intervals are constructed by extending the “Battese-Coelli” method reported by Horrace and Schmidt (1996) by allowing the technical inefficiency to be time varying and the disturbance terms to be heteroscedastic. The confidence intervals reveal the precision of technical efficiency estimates and show the deficiencies of making inferences based exclusively on point estimates. First version received: March 2000/Final version received: Oct. 2001 RID="*" ID="*"  The authors are grateful to the Economic and Social Research Council for access to their Data Archive which has provided the data for this research. We are indebted to Badi Baltagi and two anonymous referees for their helpful comments and suggestions. The usual caveat applies.     

GMM estimation of panel data models with time-varying slope coefficients

A difference/system generalized method of moments (GMM) model that imposes time-constant coefficients is common in empirical studies using panel data. However, a rejection by the Sargan–Hansen test is sometimes a serious concern for researchers. We highlight the fact that the Sargan–Hansen test for GMM estimators applied to panel data is a joint test of valid orthogonality conditions and coefficient stability over time. A possible reason for a rejection is therefore that the slope coefficients vary over time. One solution is to estimate an empirical model in which the coefficients are time specific. We apply this solution to the system GMM estimator of simple nondynamic Cobb–Douglas production functions for a selection of Swedish industries and find that relaxing the assumption of constant slope coefficients results in more satisfactory outcomes of the Sargan–Hansen test.     

Asymptotically unbiased estimation of autocovariances and autocorrelations for panel data with incidental trends

We consider the estimation of autocovariances using panel data with incidental trends under double asymptotics. The conventional autocovariance estimator suffers from a bias whose value is approximated by twice the long-run variance. We propose a bias-corrected estimator.     

Exports and growth: Granger causality analysis on OECD countries with a panel data approach

This paper investigates the possibility of Granger causality between the logarithms of real exports and real GDP in twenty-four OECD countries from 1960 to 1997. A new panel data approach is applied which is based on SUR systems and Wald tests with country specific bootstrap critical values. Two different models are used. A bivariate (GDP–exports) model and a trivariate (GDP–exports–openness) model, both without and with a linear time trend. In each case the analysis focusses on direct, one-period-ahead causality between exports and GDP. The results indicate one-way causality from exports to GDP in Belgium, Denmark, Iceland, Ireland, Italy, New Zealand, Spain and Sweden, one-way causality from GDP to exports in Austria, France, Greece, Japan, Mexico, Norway and Portugal, two-way causality between exports and growth in Canada, Finland and the Netherlands, while in the case of Australia, Korea, Luxembourg, Switzerland, the UK and the USA there is no evidence of causality in either direction.     

The effects of land market development on economic growth： an empirical analysis based on Chinese panel data, 1999-2005

Optimized land resources allocation is important for economic growth because land is one of the basic elements for economic development. And urban land resources allocation has had an increasingly important influence since the Chinese socialist market economy system was established. This paper estimates the production function of both the secondary and the tertiary industries of China＇s 31 provinces, autonomous regions and municipalities directly under the central government through an analysis of the panel data of the total output value of the secondary and the tertiary industries, invested capital, invested labor jorces and the land market-jeatured management of the above-mentioned regions during the period of 1999-2005. and examines the positive influence of the above- mentioned factors on regional economic output, This study concludes that urban economic output is positively related with the level of urban land resources market-featured management, since the rate of economic growth of those regions approximates 14. 7% under the condition of urban land market running during the period of 1999-2005.     

Spatial panel data analysis with feasible GLS techniques: An application to the Chinese real exchange rate

Recent panel data approaches stress the importance of the location interdependence. Little has been done in the Balassa-Samuelson literature accounting for spatial dependence in the panel data context that allows for spatial autocorrelation. By utilising the recently developed Kapoor et al. (2007) spatial panel feasible GLS methods, we find that the Balassa-Samuelson effect in the Chinese economy during 1985 until 2000 generally does not appear to hold. However, the black market exchange rate tends to be more compatible with the theory.     

Impact of Bt cotton on pesticide poisoning in smallholder agriculture: A panel data analysis

While substantial research on the productivity and profit effects of Bt cotton has been carried out recently, the economic evaluation of positive and negative externalities has received much less attention. Here, we focus on farmer health impacts resulting from Bt-related changes in chemical pesticide use. Previous studies have documented that Bt cotton has reduced the problem of pesticide poisoning in developing countries, but they have failed to account for unobserved heterogeneity between technology adopters and non-adopters. We use unique panel survey data from India to estimate unbiased effects and their developments over time. Bt cotton has reduced pesticide applications by 50%, with the largest reductions of 70% occurring in the most toxic types of chemicals. Results of fixed-effects Poisson models confirm that Bt has notably reduced the incidence of acute pesticide poisoning among cotton growers. These effects have become more pronounced with increasing technology adoption rates. Bt cotton now helps to avoid several million cases of pesticide poisoning in India every year, which also entails sizeable health cost savings.     

Maximum-likelihood based inference in the two-way random effects model with serially correlated time effects

The general case where the time specific effect in a two way model follows an arbitrary ARMA process has not been considered previously. We offer a straightforward maximum likelihood estimator for this case. Allowing for general ARMA processes raises the issue of model specification and we propose tests of the null hypothesis of no serial correlation as well as tests for discriminating between different specifications. A Monte-Carlo experiment evaluates the finite-sample properties of the estimators and test-statistics. We have benefitted from discussions with Pietro Balestra, Marc Nerlove and Peter Schmidt and comments from participants at the Ninth International Conference on Panel Data and the Econometric Society Eight World Congress. Financial support from HSFR, the Swedish Research Council for the Humanities and Soical Sciences is gratefully acknowledged.     

A flexible time-varying specification of the technical inefficiency effects model

The temporal pattern of technical efficiency in the technical inefficiency effects model, as modeled by Battese and Coelli (Empir Econ 20:325–332, 1995), is rather restrictive. Specifically, it a priori imposes a common pattern upon all firms in the sample, which in addition is monotonic over time. Obviously this is an undesirable implication of the model especially when there is evidence of strong firm heterogeneity and/or a long time span. To overcome this shortcoming, the present paper incorporates the Cornwell et al. (J Econom 46:185–200, 1990) flexible specification of the temporal pattern of technical efficiency into technical inefficiency effects model. The proposed formulation is then applied to the agricultural sector of the EU and US, during the period 1973–1993. The empirical result support the proposed formulation as quite different temporal patterns of technical efficiency have been found for the ten countries included in the analysis. We would like to thank an anonymous referee and an associate editor for valuable suggestions in an earlier version of the paper.     

A moving blocks empirical likelihood method for panel linear fixed effects models with serial correlations and cross-sectional dependences

To simultaneously deal with serial correlations and cross-sectional dependences for a panel linear fixed effects model, we propose a new approach based on an extended score vector and a moving blocks empirical likelihood method. Large sample properties of the proposed method are studied. Simulation results show that the new method works well under the situations of either strong or weak cross-sectional dependences, and the method performs better than the methods in Gonçalves (2011) and Vogelsang (2012). The proposed method is also applied to an application in carbon emission, and the results show that urbanization has a significant effect on carbon emission. Moreover, the effect varies in different stage of urbanization.     

Estimating elasticities of residential energy demand from panel county data using dynamic random variables models with heteroskedastic and correlated error terms

A dynamic random variables model correcting for heteroskedastic and correlated error terms over time and space and dynamic demand and using panel county data offers consistent and efficient elasticity estimates of residential electricity and natural gas demands. The model developed by Swamy [Swamy, P.A.V.B., 1974. Linear models with random coefficients. In: P. Zarembka (Eds.), Frontiers in Econometrics, Academic Press, London, pp. 143–168.] with a modification suggested by Maddala et al. [Maddala, G.S., Trost, R.P., Li, H., Joutz, F., 1997. Estimation of short-run and long-run elasticities of energy demand from panel data using shrinkage estimators. Journal of Business and Economic Statistics 15, 90–100.] uses a panel of selected California counties for the years 1983–1997 to yield elasticity estimates that differ from those obtained from more standard panel data procedures.